It's increasingly clear that with AI tools being trained based on leeching content from the internet, and the internet getting absolutely filled with AI created content, this will just create a non-sensical loop that eventually will either really slow down AI development and boost the price of custom tailored data sets, or customer-facing generative AI will implode and become really bad to the point of being mostly ignored.

they've been indexing in this manner across the board and tbh I don't quite follow the logic in some cases. for the pay-to-play search ranking shitsites (and other things that surface a lot of adspace) it makes a bit more sense, but for things like books..? why?

and I realize that this is pondering downstream effects of The No Fucks Given Company's internal development processes and things, but even so